<p>Code-switching, the fluid alternation between languages is an inherent characteristic of multilingual communication and presents both challenges and opportunities within educational contexts. This bibliometric review investigates the evolution of AI-driven research on code-switching in multilingual education from 2010 to 2025, using data extracted from the Web of Science. Employing performance analysis, science mapping, and topic modeling techniques, the study maps publication trends, evaluates influential authors and institutions, and reveals dynamic collaborative networks. Key findings underscore the centrality of translanguaging in reimagining bilingual pedagogy, the integration of large language models for improved language assessment, and the necessity of specialized, context-aware AI models. The rapid advancements in AI further amplify the importance of this research area, highlighting its critical role in shaping the future of multilingual education. The review also discusses prevailing challenges, including data quality, model adaptability, and real-world implementation, and outlines promising future directions for enhancing multilingual education through advanced AI solutions.</p>